Download

Download Full Text (783 KB)

Document Type

Poster

Description

Alcohol use disorder is a common health issue in older adults who are facing depression caused by retirement, loss of a spouse, pain, and sleep problems. The prolonged, heavy alcohol ingestion will lead to high alcohol dependency such that cessation or reduction of using alcohol causes alcohol withdrawal syndrome (AWS) in roughly 4 to 72 hours after the last drink. During the initial 8 hours, patients face anxiety, insomnia, nausea, and abdominal pain. This condition is followed by high blood pressure, increased body temperature, unusual heart rate, and confusion. If this syndrome does not receive any treatment, the patients will suffer from hallucinations, fever, seizures, and agitation. As a result, there is an essential need to predict and treat this syndrome in the initial stages. Using physiological signals, we examined the predictability of AWS over the records of patients who stayed in critical care units in Medical Information Mart for intensive care III (MIMIC-III). This dataset contains the records of 243 patients who were admitted in ICUs with the primary issue of AWS. 65% of all the patients diagnosed with mental illness as primary health concern suffered from AWS. To have a fair comparison, an equal number of records of patients without AWS are considered as the control group. We extracted nine descriptive statistical features from physiological signals and medical history of patients: average heart rate, average amount of magnesium in the blood, average body temperature, average systolic blood pressure, maximum systolic blood pressure, minimum systolic blood pressure, age, gender, and length of stay in intensive care units. The computed features were fed into 11 supervised machine learning classifiers to identify AWS conditions. The outcomes demonstrate that the Naïve Bayes classifier with an accuracy of 0.85 outperformed the others in detecting patients with AWS. This finding can help critical care physicians identify potential AWS in earlier stages and provide potential interventions before the symptoms reach a detrimental level.

Publication Date

4-2020

Disciplines

Computer Engineering | Medicine and Health Sciences

Colleges & Schools

Engineering and Computer Science

Department

Computer Science and Engineering

Faculty Advisor Name

Dr. Tanvi Banerjee

Predicting alcohol withdrawal in intensive care units


Share

COinS